Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Introduction to hybrid intelligent networks : modeling, communication, and control
Guan Z., Hu B., Shen X., Springer International Publishing, New York, NY, 2019. 292 pp. Type: Book (978-3-030021-60-3)
Date Reviewed: Feb 14 2020

Human brains are the most advanced intelligent systems to ever exist, yet also the most mysterious systems to be explored. A hybrid intelligent network is a term coined for the brain-inspired artificial intelligence (AI) that emulates particular aspects of how neurons, interconnected with synapses, operate in the brain. As such, hybrid intelligent networks usually refer to a dynamic system comprised of interconnected artificial neural networks (ANNs). Introduction to hybrid intelligent networks aims to reveal some of the science behind human brains using ANN architecture and artificial multi-agent networks.

This book begins with an informative introduction to brain-inspired intelligence and an overview of the topics and terminology to be discussed later. The rest of the chapters each focus on one aspect of hybrid intelligent networks. A typical chapter starts with an introduction and related background information. The authors then present the challenges and gaps, followed by an innovative approach supported by a rigorous formal proof and deduction. In the end, the proposed method is validated with simulation and data analysis. Due to the intensive use of formal deduction and mathematical logic processing using formal language, the majority of the book is not an easy read. However, if you enjoy the beauty of mathematical logics, deductive apparatus, and differential equations, you will feel at home. There are two major parts in the book, each with a different focus.

Part 1 (chapters 2 to 6) focuses on hybrid impulsive neural networks. Specifically, chapter 2 analyzes delayed impulsive neural networks, a particular type of hybrid intelligent network based on real-world systems featuring sudden and sharp changes occurring instantaneously, or so-called impulses. Those systems include biological neural networks, bursting rhythm models in pathology, optimal control models in economics, frequency-modulated signal processing system, flying object motions, and so on. Fundamental properties, such as the uniqueness, stability, and robustness of the equilibrium, are defined and analyzed to establish the necessary theoretical foundations.

Chapter 3 introduces a class of delayed “hybrid impulsive neural networks with interval uncertain weights.” Chapter 4 analyzes the multistability, its criteria and convergence rate, as well as the basin of attraction of equilibrium in hybrid impulsive intelligent neural networks. The application on associative memory provides an excellent model of synaptic interconnection among neurons in the brain, which is critical to learning and pattern recognition.

Chapter 5 introduces “random hybrid chaotic neural networks with nonlinear coupling and broadcast impulses.” An impulsive synchronization technique is proposed for knowledge-based sharing and intelligent automation systems. Two case studies demonstrate implementation: in image protection for better privacy and an intelligent cryptographic algorithm in the Industrial Internet of Things (IIoT).

Chapter 6 introduces memristor-based impulsive neural networks (MINN), with a memristor replacing the traditional resister in neural circuits. The multisynchronization of interconnected MINN, inspired by multitasking in the brains, is used to describe the collective behavior of neural networks. Fuzzy logic rules are used to establish the fuzzy hybrid control to achieve a positive exponential convergence rate.

Part 2 (chapters 7 to 10) studies the hybrid multi-agent network and its collective behaviors, again with thorough theoretical analyses. Specifically, chapter 7 presents a hybrid impulsive and switching control method with exponential and asymptotic stability for nonlinear systems with multi-agent networks of “arbitrary and conditioned impulsive switching.” As an example, the authors demonstrate “an impulsive control strategy ... for chaos suppression of the pendulum system.”

Chapter 8 discusses the consensus problem of linear and nonlinear multi-agent networks. A distributed hybrid impulsive control is presented, and the consensus performance measurement is investigated. Chapter 9 examines “the multiple coordination of multi-agent networks under an event-driven control paradigm” to achieve “multi-consensus without any balanced requirement on the underlying topologies.” Chapter 10 “introduces a hybrid event-time-driven asynchronous algorithm” to solve the distributed optimizing problem in sensor networks.

After being immersed in a sea of formal language symbols, readers ought to remember that a typical human brain is a universe comprised of 100 billion neurons interconnected with over 100 trillion synapses (an amount larger than the total number of atoms in the whole cosmos). The effort to study a complex subject of such scale is sacred and respectable. I am grateful to the authors who endeavor to open a window to this tiny universe.

Reviewer:  Xiangdong Che Review #: CR146890 (2005-0090)
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
Hybrid Systems (C.1.m ... )
 
 
Mathematics And Statistics (J.2 ... )
 
 
Modeling (H.5.5 ... )
 
 
Modeling And Prediction (D.4.8 ... )
 
 
Network Communications (C.2.1 ... )
 
 
Neural Nets (C.1.3 ... )
 
  more  
Would you recommend this review?
yes
no
Other reviews under "Hybrid Systems": Date
Hybrid neural network and case based reasoning system for Web user behavior clustering and classification
Zehraoui F., Kanawati R., Salotti S. International Journal of Hybrid Intelligent Systems 7(3): 171-186, 2010. Type: Article
Mar 21 2011

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy